Termsslug: agent-arena

Agent Arena

智能体竞技场

Evaluation platform comparing AI agents' task execution capabilities and costs.

Curated Intelligence

3 items · Page 1/1
Arena.ai (@arena)T1·Models
58 pts
Kimi K3.1 Pro Open Weights Top Agent Arena Open-Model Ranking

Arena.ai's weekly update reports that Moonshot released open weights for Kimi K3.1 Pro, which tops the Agent Arena open-model leaderboard with a 5.3% net gain. Tencent Hunyuan's Hy4 became the #1 Chinese model in Agent Arena at a $0.57 median cost per task and reshaped the Pareto frontier, with open weights promised soon. Google's Gemini Omni 1.2 Flash debuted at #1 in both text-to-video and image-to-video, while OpenAI's GPT Image 2.6 variants arrived at roughly twice the speed of their predecessor.

Arena.ai (@arena)T1·Models
34 pts
Tencent Hunyuan's Hy4 Full Release Nears After Strong Agent Arena Preview

Arena.ai teased the upcoming full release of Tencent Hunyuan's Hy4 without disclosing parameters, capabilities, or a date. Its cited preview results show Hy4 preview ranking #2 among open models (#10 overall) across 14.5K+ real-world agent sessions, at a median task cost of $0.26 versus $0.80 for Kimi K3 (Max), placing it on the Agent Arena Pareto frontier.

Latent SpaceT1·Models
78 pts
OpenAI Launches GPT-6.1 Sol: Cutting Costs While Boosting Benchmarks

OpenAI has launched GPT-6.1 Sol, priced at just 20% of its predecessor Astra ($2/$10 per M tokens), while achieving significant score improvements on key coding and automation benchmarks like DeepSWE and AutomationBench. Positioned as 'good, cheap AND fast,' the model entered the top 5 in Agent Arena with extremely low median cost per task, marking a major breakthrough in the cost-performance frontier for large models.